Bloomberg reports that British Columbia sued OpenAI in California, alleging it could have used ChatGPT logs to warn police and avert a mass shooting. A post sharing the report says OpenAI explained that it banned a ChatGPT account in June 2025 but did not refer it to police because the activity did not meet its then-threshold for “credible and imminent” planning. The post also quotes OpenAI telling Canadian officials on February 26, 2026, that under its revised referral protocol, it would refer that account to law enforcement if it were discovered that day.